GIS Current Transformer Intelligence Solution: Mechanical Fault Early Warning System Based on Vibration-Current Coupling

Core Problem:​​ In seismic-prone regions and aging GIS substations, the mechanical structures (e.g., fasteners, insulation supports) of Current Transformers (CTs) are susceptible to damage from continuous vibration or sudden impacts. This can lead to hidden faults like loosening, detachment, or displacement, ultimately causing insulation degradation or sudden CT failure, threatening grid reliability. Traditional outage-based inspection methods are inefficient and costly.

​Innovative Solution:​​ Integrates vibration and current dual-parameter monitoring, utilizing an AI engine to achieve early warning and intelligent diagnosis of CT mechanical faults.

​Core Technological Implementation​

1. ​Multi-Parameter Collaborative Sensing:​​
• ​High-Frequency Vibration Monitoring:​​ Deploy wide-band piezoelectric accelerometers (5Hz-10kHz) on critical CT components (flanges, supports) to accurately capture abnormal structural vibration signals induced by mechanical loosening, part displacement, insulation degradation, or external vibrations (seismic waves).
• ​Transient Inrush Current Capture:​​ Employ passive Rogowski Coils for non-intrusive, real-time monitoring of CT primary-side switching operation current waveforms. Combined with circuit breaker operation signals, it precisely identifies switching events and analyzes inrush characteristics and their impact forces on the CT mechanical structure.
2. ​AI-Driven Edge Intelligence Diagnostic Engine:​​
• Employs a ruggedized edge computing module (wide-temperature, shock-resistant) installed locally on the device for real-time processing of vibration and current waveform data.
• Core operation utilizes a proprietary ​1D-CNN (1D Convolutional Neural Network)​​ intelligent diagnostic model:
• ​Input:​​ Vibration acceleration time-frequency features (FFT analysis) + Switching inrush waveform characteristics.
• ​Output:​​ Accurately identifies typical mechanical fault modes ("bolt loosening," "insulation support displacement," "mechanical resonance") with a diagnostic accuracy rate of 92%.
• Possesses ​​"adaptive learning" capability​ to adapt to different CT structural and background vibration characteristics across various substations.
3. ​Efficient Local Warning & Communication:​​
• ​Tiered Warning Mechanism:​​ Upon detection of suspected fault signatures, the edge engine immediately generates warning/alarm signals (e.g., Warning, Severe, Critical).
• ​Simplified Wireless Transmission:​​ Encrypted transmission of key alarm signals (not raw data) to the substation's local HMI platform via LoRa LPWAN technology, significantly reducing communication load and latency.
• ​Local HMI Display:​​ Real-time map-based display showing the faulted CT number, fault type, alarm level, and recommended actions.

​Target Application Scenarios​

1. ​GIS Substations in High-Seismic Areas:​​
• Early warning of CT displacement or structural damage triggered by seismic aftershocks, preventing secondary faults.
• Continuous monitoring of chronic equipment damage caused by long-term minor geological activity.
2. ​Aging GIS Substation Retrofits & Upgrades:​​
• ​Outage-Free Deployment:​​ Simple sensor installation requiring no modification of gas chamber structure, ensuring airtight integrity. Particularly suitable for older stations with limited outage windows.
• ​Cost-Effective Incremental Retrofit:​​ Leverages wireless tech and edge computing, eliminating the need for extensive cabling or new backend systems, resulting in high retrofit ROI.
3. ​Critical Hub & High-Load Substations:​​ Prevention of protective relay misoperation/failure risks and large-scale blackouts caused by sudden CT failure.

​Core Value & Advantages​

• ​Early Major Risk Warning:​​ Effectively predicts typical mechanical faults ​7+ days in advance, providing ample time for proactive intervention.
• ​Significant Reduction in Unplanned Outages:​​ Cuts unplanned outages caused by sudden CT failure by ​over 60%​, significantly enhancing grid availability and customer satisfaction.
• ​Dual Safety & Cost Benefits:​​ Prevents protection system malfunctions, arc faults, and cascading equipment damage within the substation resulting from CT failure.
• ​New Smart O&M Paradigm:​​ Shifts from "periodic maintenance" to ​​"predictive maintenance"​, greatly optimizing spare parts management and workforce scheduling.
• ​Design Suited for Strong Seismic Zones:​​ Passive sensors + Edge computing + Wireless transmission – no long cabling, resilient to strong earthquakes.
• ​Optimal Cost for Aging Station Retrofits:​​ Lightweight solution, independent of large substation monitoring systems; ensures rapid deployment and payback.